Report: Honda to assemble Fit hybrids in Thailand
Vauxhall Ampera offered through Zipcar car sharing scheme

Simulation study finds “Millerizing” the Scuderi split-cycle engine could enable 50% engine downsizing with retention of power and torque, along with reduced fuel consumption

Simulation study finds Miller version of Scuderi engine can deliver twice the BMEP as naturally aspirated version, theoretically enabling 50% downsizing while maintaining performance. Source: Scuderi. Click to enlarge.

A simulation study by researchers at the Southwest Research Institute (SwRI) has found that applying turbocharging and the Miller cycle to the Scuderi split-cycle engine (SSCE) could support engine downsizing of 50% relative to a naturally-aspirated version of the Scuderi engine along with a reduction in fuel consumption while maintaining power and torque. (Earlier post.)

The Miller cycle—an approach to improving the thermal efficiency of a combustion engine—uses a higher expansion ratio than compression ratio (i.e., over-expansion). For conventional reciprocating ICEs, this involves early or late closing of the intake valves, resulting in a smaller effective compression stroke; combustion and expansion proceed normally. In a naturally aspirated conventional engine, this would result in a loss of BMEP, but an increase in thermal efficiency.

Applying the Miller cycle to the Scuderi engine (right) entails downsizing of the compression cylinder. Source: Scuderi. Click to enlarge.

In practice, David Branyon and Dean Simpson note in their paper, the Miller cycle is usually applied to turbocharged engines, where BMEP can be maintained via the higher boost level offsetting the loss in trapped volume.

The Scuderi split-cycle design divides the four strokes of the conventional combustion cycle over two paired cylinders: one intake/compression cylinder and one power/exhaust cylinder, connected by a crossover port. The two cylinders are out of phase by 20 degrees crank angle, and the charge mass is transferred from the compression cylinder to the expansion cylinder during the period between their respective TDCs.

The fundamental approach to this work was to accomplish over-expansion of the combustion gases through differential sizing of the combustion and expansion cylinder, rather than through variation of intake valve closing time.

On the split-cycle engine, the definition of effective compression and expansion ratios are not clearly obvious. The geometric ratios are very high for each cylinder/piston combination, but near top dead center (TDC), the crossover valves are open, notionally adding the crossover passage volume to the effective clearance volume to each cylinder.

For fixed cylinder clearance volumes, the displacement volumes of the SSCE [Scuderi Split Cycle engine] cylinders will effectively control the compression and expansion ratios with regards to the Miller cycle.

—Branyon and Simpson

For the study, the SwRI team applied the Miller cycle to a spark-ignited, stoichiometric, gasoline-fueled version of the SSCE with downsized compressor cylinder in 1D cycle simulation software. They then used a simple, thermodynamic model of a turbo to apply a fixed boost level. Among several positive interactions between the split-cycle design and the Miller cycle they found were:

  • Reduction of compression cylinder displacement allowed Miller cycle operation while still closing the intake valve at an optimum trapped mass condition. This results in more favorable pumping work than the Miller cycle applied in a conventional ICE due to the avoidance of intake valve closing during a period of high piston velocity.

  • The high turbulence and resulting fast combustion and late fuel addition provides a natural knock avoidance characteristic that allows utilization of higher boost levels than are typically achievable with stoichiometric, spark-ignited engines.

They found that application of the Miller cycle led to large improvements in both BMEP and fuel consumption (BSFC) at full load operating points at 1400 and 4000 rpm; performance gains at 1400 rpm were greater than those at 4000 rpm. At 1400 rpm, BMEP increased to more than 21 from 15 starting at 1.7 bar absolute boost all the way to maximum boost analyzed. At the same time, fuel consumption dropped from 250 g/kWh to less than 240 g/kWh.

After the analysis of 4000 and 1400 rpm maximum load performance was completed, consideration was made of the optimum configuration for a multi-speed engine running across this speed range and lower. Primarily, this mean constraining the compressor stroke to be the same at the two conditions.

...Selecting the points at 55 mm compressor stroke from each speed analysis...results in the achievement of just under 19 bar BMEP at 4000 rpm and about 19.5 bar BMEP at 1400 rpm. This is roughly twice the nominal naturally-aspirated engine BMEP targets.

...As the Miller cycle SSCE engine is able to produce twice the BMEP and hence specific power compared to the naturally-aspirated version, it means that the engine could be downsized by 50% for the same maximum power, torque and vehicle performance...the increase in full load BMEP and the downsizing that is facilitated by that high BMEP pays large dividends in terms of light-load fuel consumption, which makes up the bulk of light vehicle drive cycles.

—Branyon and Simpson

Future work on the Millerized Scuderi engine will use real turbocharger maps with a real turbocharger model in the cycle simulation to verify that the results can be obtained with real hardware.


  • David Branyon and Dean Simpson (2012) Miller Cycle Application to the Scuderi Split Cycle Engine (by Downsizing the Compressor Cylinder) (SAE 2012-01-0419)



Well, finally, in the notso near future, they will use real hardware . . . .

Work will “use real turbocharger maps with a real turbocharger model in the cycle simulation”.

And they say this will “verify that the results can be obtained with real hardware.” ?

Maybe it will sidestep the fact that with overexpansion there is little or nothing left to drive a turbocharger.


Use electric driven turbochargers?

If this can be done at an affordable cost, could it become part of a more efficient and lighter on-board genset for 100 mpg HEVs and 200 mpge PHEVs?


There's plenty to drive the turbocharger even with overexpansion; when you look at the actual pressure, temperature and γ of the combustion gases, it's instantly obvious that it takes more than a 2:1 overexpansion to use up the available expansion energy.  Furthermore, once the turbo begins operating, the available exhaust energy increases.

240 g/kWh still isn't all that impressive.

Nick Lyons

Simulation blah blah blah


They're just making noise to try and get more funding.


The figure I'm finding for the Prius engine is 220 g/kWh.

It wouldn't be hard to eliminate the flow losses from the conventional Atkinson, such as using over-long intake runners to delay the point of full cylinder filling until well after BDC.


Paid dreamers


I work more then them and i do not receive subsidies nor salary.


The combustion chamber will survive how?


Turbo Scuderi:  240 g/kWh

Mahle jet ignition:  190 g/kWh.

I always thought that Scuderi's use of a cold crossover passage was a mistake for thermal efficiency, and this now seems to be borne out.  Scuderi should have been trying to keep the compressed charge hot, as well as minimizing losses due to pressure drops.

The comments to this entry are closed.